Colonoscopy plays a crucial role in the diagnosis and prognosis of various gastrointestinal diseases. Due to the challenges of collecting large-scale high-quality ground truth annotations for colonoscopy images, and more generally medical images, we explore using self-supervised features from vision transformers in three challenging tasks for colonoscopy images. Our results indicate that image-level features learned from DINO models achieve image classification performance comparable to fully supervised models, and patch-level features contain rich semantic information for object detection. Furthermore, we demonstrate that self-supervised features combined with unsupervised segmentation can be used to discover multiple clinically relevant structures in a fully unsupervised manner, demonstrating the tremendous potential of applying these methods in medical image analysis.
A model's interpretability is essential to many practical applications such as clinical decision support systems. In this paper, a novel interpretable machine learning method is presented, which can model the relationship between input variables and responses in humanly understandable rules. The method is built by applying tropical geometry to fuzzy inference systems, wherein variable encoding functions and salient rules can be discovered by supervised learning. Experiments using synthetic datasets were conducted to investigate the performance and capacity of the proposed algorithm in classification and rule discovery. Furthermore, the proposed method was applied to a clinical application that identified heart failure patients that would benefit from advanced therapies such as heart transplant or durable mechanical circulatory support. Experimental results show that the proposed network achieved great performance on the classification tasks. In addition to learning humanly understandable rules from the dataset, existing fuzzy domain knowledge can be easily transferred into the network and used to facilitate model training. From our results, the proposed model and the ability of learning existing domain knowledge can significantly improve the model generalizability. The characteristics of the proposed network make it promising in applications requiring model reliability and justification.
Optical colonoscopy is an essential diagnostic and prognostic tool for many gastrointestinal diseases including cancer screening and staging, intestinal bleeding, diarrhea, abdominal symptom evaluation, and inflammatory bowel disease assessment. However, the evaluation, classification, and quantification of findings on colonoscopy are subject to inter-observer variation. Automated assessment of colonoscopy is of interest considering the subjectivity present in qualitative human interpretations of colonoscopy findings. Localization of the camera is an essential element to consider when inferring the meaning and context of findings for diseases evaluated by colonoscopy. In this study, we proposed a camera localization system to estimate the approximate anatomic location of the camera and classify the anatomical colon segment the camera is in. The camera localization system starts with non-informative frame detection to remove frames without camera motion information. Then a self-training end-to-end convolutional neural network was built to estimate the camera motion. With the estimated camera motion, the camera trajectory can be derived, and the location index can be calculated. Based on the estimated location index, anatomical colon segment classification was performed by building the colon template. The algorithm was trained and validated using colonoscopy videos collected from routine clinical practice. From our results, the average accuracy of the classification is 0.759, which is substantially higher than the performance of using the location index built from other methods.